Impact on ambulatory palliative appointment lead times by standardizing scheduling protocols based on cognitive computing models.

Abstract

62 Background: Although palliative care is available across the United States, access continues to remain variable. In Texas, only 52% of hospital systems have palliative services. As a result, optimizing existing systems is critical to providing timely services to this vulnerable population. The Supportive Care Services Clinic at University Health in San Antonio, TX provides ambulatory palliative services to patients including physical and psychosocial symptom management and care coordination. For the period April 1, 2023 to June 30, 2023 average lead times for new palliative patient visits (NPPV) and hospital consult discharge visits (HCDV) was 53 days and 39 days respectively. The aim of this project was to decrease the lead times for NPPV and HCDV through standardization of scheduling criteria. We hypothesized that by applying a scheduling protocol utilizing cognitive computing models (CCM) calculating readmission risk scores to prioritize referrals, we could reduce appointment lead times. Methods: During the implementation phase, we developed referral triage criteria to prioritize patients on a scheduling wait list, including patients who were at higher risk of hospitalization or rapid decline based on CCM readmission risk scores calculated within the electronic medical record. Patients were assigned to next available in template. If a cancelation occurred, patients were pulled forward and vacancy was filled based on priority score. Results: For the initial post-implementation period between September 1, 2023 and November 30, 2023 NPPV average lead time was 47 days which was a reduction of 11% from baseline. The average lead time for HCDV was 33 days which was a reduction of 15% from baseline. In a second follow-up period from February 1, 2024 to April 30, 2024, the average lead time of NPPV was 22 days, a 58% decrease from baseline. The average lead time for HCSV was 24 days which was a significant reduction (p-value = 0.011) of 38% from the pre-implementation baseline. Conclusions: Utilization of CCM risk scores to triage NPPV and HCDV significantly reduced the lead times and resulted in more timely services for critical patients. We also noted increased template utilization. Future projects aim to study the effects of scheduling protocols on template utilization, cancellation rates, no-show rates, readmission rates and other health care utilization. Palliative appointment lead times. April-June 2023 Sept-Nov 2023 Feb-Apr 2024 Total Pt Seen 292 364 534 Avg Lead Time (d) 41 38 29 New Patient New Patient New Patient Completed 66 82 74 Avg Lead Time (d) 53 47 22 Hospital Discharge Hospital Discharge Hospital Discharge Completed 27 28 25 Avg Lead Time (d) 39 33 24